Imaging systems of this type are known as "line cameras" and are particularly used for inspection tasks in the industrial manufacturing field. They include a unidimensional detector array on which the objects are imaged by means of a conventional lens system. Unidimensional detector arrays are substantially less expensive than bidimensional matrices and are available as a single chip accommodating a relatively large number of individual elements. However, to generate a two-dimensional image, a relative motion between object image and array is necessary, this motion being accomplished either by mechanically moving mirror systems located in the imaging beam path or by a movement of the object itself. These motion cycles limit the maximum achievable image frequencies and thus the usability of these systems. In addition, the equipment necessary for mechanically scanning the image in one direction of the coordinates is costly and voluminous.
In the journal entitled "Optical Spectra", May 1980, pages 51 to 52, an image sensor is described which includes a bidimensional CCD (charge-coupled device) array. More specifically, the sensor includes a photoelectric solid-state matrix of the type known as a capacitive bucket-brigade device having a light-sensitive layer upon which is placed a thin fiber-optic image-conducting plate. The front end of this plate is covered by a still thinner glass plate into which light is coupled laterally via glass fibers for illuminating the object to be imaged.
The sensor is applied directly to the object to be imaged and provides an electronic contact print thereof.
Such a sensor, which dispenses with the need for an optical lens system, possesses certain advantages over conventional imaging systems because it permits a very compact design and may thus be used, for example, for the examination of inaccessible locations of the body or machine parts in the manner of an endoscope or technoscope. Its disadvantages are among others that it uses a bidimensional semiconductor matrix for image conversion. At least at the present time, such components still have a low resolution and are relatively expensive, since the production output of perfect components is rather poor. Also, it is particularly in medical applications that getting electrical supply and signal lines to the object to be examined is problematic.
Therefore, known image sensors used for endoscopy include a bundle of fiber-optic conductors which transmit the image of the object to a conventional tube camera, the image being generated remotely by means of an optical lens system on the end face of the fiber-optic bundle.